Competition in the banking industry, is it beneficial? Evidence from MENA region

Abstract

Using a sample of 197 commercial banks operating in the MENA region, this paper examines the impact of competition on bank profitability and risk-taking, during the period 2011–2018. To assess competition, we have used both market structure-based and non-market structure-based indicators. Results indicate that competition would impact differently banks’ profitability and risk-taking. On the one hand, an increase in competition seems to hurt banks’ net—interest margin and loan quality, but on the other hand, the competition seems to enhance banks’ ROA and ROE. In line with existing literature, we suggest that competition reduces interest-based income, and increases bank fragility, but it seems to enhance the development of new non-interest-based activities, improving hence the overall banks’ profitability. We argue that our findings would contribute to the current debate on the effectiveness of the deregulation reform initiated in several MENA countries and help to resolve the trading-off between competition and bank stability.

Appendix 1: Lerner index calculation

Price can be calculated by computing the ratio of the total revenue to the total output. Marginal cost can be determined by employing a trans-log cost function. The homogenous price index is used in this equation so that the price of the input is normalized by one of the input prices. As Yuanita, we have used the following the formula for trans-log cost function:

$$\begin{aligned} {\text{LnTC}} & = \alpha_{0} + \alpha_{1} {\text{Ln}}Y + \frac{1}{2}\alpha_{2} ( {\text{Ln}}Y )^{2} + \sum\limits_{J = 1}^{3} {\beta_{j} } w_{j} + \sum\limits_{J = 1}^{3} {\frac{1}{2}} \beta_{j} {\text{Ln}}w_{j}^{2} + \sum\limits_{j = 1}^{3} {\sum\limits_{k = 1}^{3} {\beta_{jk} {\text{Ln}}w_{j} w_{k} } } + \sum\limits_{j = 1}^{3} {\lambda_{j} {\text{Ln}}Y{\text{Ln}}w_{j} } \\ & \quad + \sum\limits_{j = 1}^{3} {\rho_{j} } {\text{trendLn}}w_{j} + \eta_{0} ( {\text{Ln}}Y*{\text{trend)}} + \eta_{1} {\text{trend}} + \frac{1}{2}\eta_{2} {\text{trend}}^{2} + \varepsilon \\ \end{aligned}$$

where TC is the total cost, y is the total loan, W1 is the price of labor (the ratio of personnel cost to the total assets), W2 is the price of physical capital (the ratio of all other costs excluding interest, personnel, and impairment costs to the total assets), W3 is the price of borrowed funds (the ratio of interest cost-to-third-party funds). W4 is the price of capital charged for assets deterioration (the ratio of impairment costs to the total assets). Total cost and input prices are normalized by W4 to have a homogenous price standard.

The trans-log cost function is computed using generalized least square. The regression result is used to calculate marginal costs by employing the first derivative of the cost function for the total loan as shown below:

$${\text{mc}} = \frac{\text{TC}}{y} (\alpha_{1} + \alpha_{2} {\text{Ln}}Y + \sum\limits_{j = 1}^{3} {\gamma_{j} {\text{Ln}}w_{j} } + \mu_{0} {\text{trend)}}$$